Whether to delegate

You shouldn’t necessarily do things that you’re good at. In economics, this idea is known as comparative advantage. Delegating may free up your time to do something more profitable. It might be to a country’s advantage to import something that they could produce cheaper domestically. Importing one thing might free up resources to export another thing that’s more valuable.

Comparative advantage is often illustrated by a hypothetical lawyer and an assistant. A lawyer who can type very quickly is still better off hiring someone else to do the typing because he can make much more per hour practicing law. If he could type twice as fast as an assistant, and he could earn more than twice as much practicing law as it costs to hire an assistant, he makes money by delegating.

This illustration makes sense at one level, but it also sounds a little quaint. In fact lawyers do quite a bit of typing. That’s explained by another economic idea: transaction costs. It costs time to recruit and hire an assistant. And once you have an assistant, it takes time to explain what you want done, time to wait for the work to come back, time to review the work, etc.

Highly paid executives type their own emails, at least some of the time, because it’s not worth the transaction costs to have someone else do it. But for a larger task, say typing up hundreds of handwritten pages, it’s worth paying the transaction costs to get someone else to do the typing.

Most advice on delegation is simplistic. It ignores transaction costs, and has a naive view of opportunity costs. It says that if you make $50 an hour, you should delegate anything you can hire done for $40 an hour since the opportunity cost of doing the $40 an hour task rather than delegating it is $10 an hour. But things are more subtle than that.

Opportunity costs only apply if you’re turning down an opportunity. If you stop doing $50 an hour work to do $40 an hour work, then you’re losing $10 an hour compared to what you could earn (ignoring the transaction costs of delegating). But if you don’t have $50 an hour work to do, if you’re otherwise idle, then delegating $40 an hour work is costing you $40 an hour, not saving you $10 per hour.

People are not machines. If you have an idle machine, give it work to do. And if two machines could do the same work, use the one that can do the work the cheapest. But people are more complicated. We like some kinds of work better than others, we learn, and we need time to rest.

Suppose you enjoy doing work that you could delegate for $40 an hour. You find it refreshing. There’s no opportunity cost in doing it yourself if the time to do it comes out of time you would have spent on a hobby.

Suppose you don’t enjoy doing work that you could delegate, but there’s something you could learn from doing it. In that case, there may be an opportunity benefit as well as an opportunity cost: learning something new may create opportunities in the future.

The previous two paragraphs account for enjoyment and learning, but not rest. If you don’t have $50 an hour work to do, doing $40 an hour work is only one alternative. Another alternative is to do nothing, which is very valuable in ways that are hard to quantify. And even work you enjoy may take energy away from other work.

Managing energy is more important than managing time. Energy is what gets things done, and time is only a crude surrogate for energy. Instead of only looking at what you could earn per hour versus what you could hire someone else for per hour, consider the energy it would take you to do something versus the energy it would free to delegate it.

If something saps your energy and puts you in a bad mood, delegate it even if you have to pay someone more to do it than it would cost you do to yourself. And if something gives you energy, maybe you should do it yourself even if someone else could do it cheaper.

Finally, note that energy isn’t the same as pleasure, though they often go hand in hand. Some activities are enjoyable but draining, and some are not enjoyable but invigorating. For example, I enjoy teaching, but it takes a lot out of me. And most people don’t enjoy exercise that much even though it gives them energy.

18 thoughts on “Whether to delegate

  1. Josh: Agreed! A lot of economic abstractions make sense in a certain context and at a certain scale but break down outside of that.

  2. Forrester’s System Dynamics builds ordinary differential equation models, sometimes including things such as practice effects, transaction costs, burnout.

    If you have a particular rule of thumb, more complicated than “delegate tasks that you can have done for cheaper than your hourly rate”, then you can probably reverse-engineer the rule of thumb into a model – the simplest model for which that rule of thumb is an optimal policy.

    I’d love to see more of those “implicit models” behind rules of thumb.

  3. Just a minor niggle: machines are not that simple either. As a human assistant has hiring costs, a machine has evaluation costs. Perhaps analogous to payroll taxes and employer-provided insurance, machines have warranties and service contracts. Cloud computing and mainframe capacity on demand are somewhat analogous to temporary workers (rapid response to increased work). Idle machines also have costs (storage and depreciation–and some have higher technological depreciation or higher restart costs).

    Wear and tear and productivity of a machine can vary by workload (which has some similarity to effects of human liking). While a machine might not have much learning potential, configuration (including adding components) and user familiarity have similar effects.

    (This is just an obvious quibble.)

  4. Instead of viewing the relevant economic abstractions as ‘breaking down’, you could view what you’ve done in this post as more fully applying those abstractions.

    Thanks for the post!

  5. Delegation also builds the skill set of junior colleagues, offers opportunities for advancement, and increases the pool of stakeholders who feel they have ownership of a project in an organization. If you listen to the people you delegate to, you can sometimes learn something from alternate points of view. Economic models seem to miss important non-quantifiable aspects of decision-making.

  6. I agree with David. It reminded me of one of your older posts about bicycle skills. The transaction cost of delegation may be high at first, but decrease with each act, as junior staff gain experience and skill.

    Delegation also increases the resilience of an organization. I like to think we should all be able to suffer being hit by a bus, without our work unit failing completely.

  7. Experts are always advising entrepreneurs to delegate as much as possible, but like you mentioned, “if you don’t have $50 an hour work to do, if you’re otherwise idle, then delegating $40 an hour work is costing you $40 an hour, not saving you $10 per hour.”

    On the business side, early on I think entrepreneurs would be better served looking for ways to systematize operations vs. simply delegating. On the personal side, I would agree that managing time, especially when overworked, is pretty pointless unless you manage energy first.

  8. Anselmo Pitombeira

    I think there is a flaw in your reasoning, and it has to do with the definition of opportunity costs. Monetary costs are only a part of opportunity costs, but there are also implicit costs, which are non monetary. The definition of opportunity costs by Greg Mankiw is: whatever must be given up to obtain some item. Using your example, when the lawyer decides to do the typing, her opportunity cost is not $10. It is all other things she could alternatively do, including taking a nap or doing nothing. The usefullness of the opportunity cost concept relies on the ideia that someone has always to take account in a decision what she is giving up when a choice is made.

  9. I agree with a lot of this (the second point is that the cost to consider is your marginal cost, not your average cost; it’s pretty standard Econ 101).

    However, do you think most people over-delegate? I think most under-delegate.

    Organizations often over-delegate internally, but that is rarely justified in the way you say it, rather it is about the efficiency of scales that come if you have everyone buy a plane ticket through a centralized system rather than individually on expedia (curiously enough, the same sort of organization is sometimes justified on the basis of its inefficiency: “it creates jobs”).

  10. I don’t see any contradiction with the basic principle of comparative advantage. You want to think in terms of ‘energy’ instead of time, fine. But the principle still applies. You spend energy in the most productive activity and delegate if the energy balance from doing so is positive. You are only introducing more detail into the model, which is fine, but you are not disproving any basic principle…

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>